|Home | About | Journals | Submit | Contact Us | Français|
Human cytomegalovirus DNA is packaged in virions without histones but associates with histones upon reaching the nucleus of an infected cell. Since transcription is modulated by the interplay of histone modifications, we used chromatin immunoprecipitation to detect acetylation and methylation of histone H3 at viral promoters at different times during the viral replication cycle. Histone H3 at immediate-early promoters is acetylated at the start of infection, while it is initially methylated at early and late promoters. Acetylation at immediate-early promoters is dynamic, with a high level of activating modifications at 3 and 6 h postinfection (hpi), followed by a marked reduction at 12 hpi. All viral promoters, as well as nonpromoter regions, are modified with activating acetylations at 24 to 72 hpi. The transient reduction in histone H3 acetylation at the major immediate-early promoter depends on the cis-repressive sequence to which the UL122-coded IE2 protein binds. A mutant virus lacking this element exhibited decreased IE2 binding at the major immediate-early promoter and failed to show reduced acetylation of histone H3 residing at this promoter at 12 hpi. Our results demonstrate that cytomegalovirus chromatin undergoes dynamic, promoter-specific histone modifications early in the infectious cycle, after which the entire chromosome becomes highly acetylated.
A high percentage of the population has been infected with human cytomegalovirus (HCMV), the prototypical member of the Betaherpesviridae family, and carries the virus in a latent state. For most immunocompetent people, infection does not cause serious disease. In contrast, primary infection or reactivation in immunocompromised individuals can be associated with serious pathology or mortality (42).
HCMV genes, like those of other herpesviruses, are expressed sequentially upon infection of permissive cells (60). The first genes expressed are the immediate-early genes. The major immediate-early promoter (MIEP) controls the production of mRNAs encoding the viral IE1 and IE2 proteins which interact with a variety of cellular proteins to regulate subsequent viral gene expression (7, 15, 17, 34). IE2 also binds DNA directly to modulate transcription (1, 8, 29, 30). The next group to be transcribed are the early genes, some of which are required for viral DNA replication (58). Once viral DNA replication occurs, it leads to the transcription of the late genes; this group includes most of the structural proteins that make up the capsid and the tegument (16).
The transcriptional regulation of cellular genes is controlled at the level of chromatin by the interplay among numerous modifications of histones (61). For instance, acetylation of certain lysine residues in histones H4 and H3 is associated with active transcription (9, 41). While histone H4 is acetylated at lysines 5, 8, 12, and 16 (67), it appears that lysine 16 acetylation may be the founding modification event (57). In the case of histone H3, acetylation at lysine 9 and 14 lead to active transcription (33, 47). Besides hosting an activation marker, lysine 9 of histone H3 can initiate transcription silencing when it is methylated (45). This methylation recruits heterochromatin protein 1 (HP1), which causes chromatin to be remodeled into a closed inactive state (3, 27).
Histone modifications and the presence of cellular and viral regulatory proteins at the HCMV MIEP have been studied under a variety of conditions. MIEP activity correlated with the presence of acetylated histone H4 versus HP1 at the promoter in infected human teratocarcinoma (T2) cells (44). T2 cells are nonpermissive for HCMV infection but become permissive after differentiation with retinoic acid. The MIEP is associated with higher levels of HP1 in undifferentiated than in differentiated T2 cells, and with higher levels of acetylated histone H4 after differentiation compared to before differentiation. A similar correlation between these markers and transcriptional activity was observed for the MIEP in nonpermissive CD34+ cells, a cell population that hosts latent virus, versus differentiated, permissive dendritic cells (51). Experiments performed with THP-1 cells, a monocytic cell line, also revealed a correlation between the activity of viral promoters and chromatin markers of transcriptional activity (20). In undifferentiated, nonpermissive THP-1 cells viral promoters were associated with a lower level of acetylated H4 and more dimethylated H3, as expected for inactive chromatin. In contrast, after infection of differentiated, permissive THP-1 cells viral promoters contained more acetylated H4 and less dimethylated H3, consistent with active chromatin. The IE1 and IE2 proteins induce activating chromatin markers at early promoters in undifferentiated THP-1 cells. The IE2 protein has been shown to be present at the MIEP, as well as the UL112 and UL4 early promoters within transfected cells (48), and transfection experiments have further demonstrated that IE2 interacts with histone deacetylase (HDAC) and methyltransferase at the MIEP (50). Finally, IE2 has been shown to bind and influence histone modifications at the MIEP within infected cells (50).
In order to better understand the relationship between histone modification and HCMV gene transcription, we examined the changes in acetylation and methylation at multiple HCMV promoters as the virus progressed through its replication cycle in fibroblasts. Our results demonstrate that histones at immediate-early promoters are marked by activating acetylations at the start of infection, many hours before the histones at early and late promoters are acetylated. Acetylation at immediate-early promoters is dynamic, with a high level of activating modifications at 3 and 6 hpi, followed by a marked reduction at 12 hpi. The transient reduction in histone acetylation at the MIEP does not occur in a mutant virus lacking the cis-repression sequence (CRS) to which the IE2 protein binds. Late after infection, both promoter and nonpromoter regions of the genome contain acetylated histones.
Human MRC-5 embryonic lung cells (American Type Culture Collection) were grown in Dulbecco modified Eagle medium supplemented with 10% fetal calf serum. The wild-type AD169 strain of HCMV used for the present study, BADwt, was derived from an infectious BAC clone of the viral genome (66). An AD169 derivative containing a mutated CRS, termed BADsubUL122crs−, was generated by using the galK selection and a counterselection “recombining” protocol (63) as applied to BAC-cloned HCMV previously (43, 63). The forward and reverse primers used to insert the galK expression cassette into the MIEP were 5′-CGCAAATGGGCGGTAGGCGTGTACGGTGGGAGGTCTATATAAGCAGAGCTCcctgttgacaattaatcatcggca-3′ and 5′-GTCTTCTATGGAGGTCAAAACAGCGTGGATGGCGTCT CCAGGCGATCTGACtcagcacgttcctgctcctt-3′ (capital letters indicate the viral sequence). The oligonucleotides used to generate the mutant CRS were 5′-GTA GGCGTGTACGGTGGGAGGTCTATATAAGCAGAGCTCaggTAGTGAACCG TCAGATCGCCTGGAGACGCCATCCACGCTGTTTTGACC-3′ and 5′-GGTCAAAACAGCGTGGATGGCGTCTCCAGGCGATCTGACGGTTCACTAcctGAGCTCTGCTTATATAGACCTCCCACCGTACACGCCTAC-3′ (lowercase letters designate mutation within the CRS).
Infection of MRC-5 cells was performed at a multiplicity of 3 PFU/cell. Virus was allowed to adsorb to cells for 60 min. Residual, unattached virus was then washed off the cell monolayer with phosphate-buffered saline (PBS), and medium supplemented with 10% newborn calf serum was added to the cells. Virus particles produced in MRC-5 cells were partially purified by centrifugation through a 20% sorbitol cushion, resuspended in Dulbecco modified Eagle medium supplemented with 10% newborn calf serum, and stored at −80°C until used.
HDACs were inhibited with sodium butyrate (10 mM; Upstate Biotechnology) or trichostatin A (TSA; 300 nM; Sigma). The HCMV DNA polymerase was inhibited by treating the cells with phosphonoacetic acid (PAA) (1.2 mM; Sigma), and translation was blocked by treatment with cycloheximide (100 μg/ml; Sigma).
At 3, 6, 12, 24, 48, and 72 h postinfection (hpi) nuclear proteins were cross-linked to DNA by the addition of 1% formaldehyde to culture medium supplemented with 10% newborn calf serum for 10 min, after which the reaction was quenched by adding glycine to a final concentration of 0.125 M, followed by incubation for an additional 5 min at room temperature. The cells were washed twice with ice-cold PBS, scraped, collected into ice-cold PBS supplemented with a protease inhibitor cocktail (Roche), centrifuged (700 × g for 5 min), and then resuspended in lysis buffer (10 mM Tris-HCl [pH 8.0], 100 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, 0.1% sodium deoxycholate, 0.5% N-lauroylsarcosine, protease inhibitors). Cells were disrupted, and DNA was fragmented by sonication (six sets of 25 pulses using a Fisher Scientific 550 Sonic Dismembrator). Cellular debris was removed by centrifugation (14,000 rpm for 10 min), and a fraction of the lysate was kept to be used as input samples and to quantify the amount of viral DNA (7800HT sequence detector system; Applied Biosystems). An aliquot of lysate containing 3 μg of viral DNA was diluted 1:10 in dilution buffer (0.01% sodium dodecyl sulfate [SDS], 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl [pH 8.0], 167 mM NaCl, protease inhibitors) and then incubated overnight at 4°C with rotation after the addition of 5 μg of an antibody specific for histone 3 (catalog no. 07-690; Upstate Biotechnology), acetyl K9H3 (catalog no. 07-352; Upstate Biotechnology), acetyl K14H3 (catalog no. 07-353; Upstate Biotechnology), dimethyl K9H3 (catalog no. 07-441; Upstate Biotechnology), IE2 (5A2 monoclonal antibody produced by immunization with a fusion protein containing the C-terminal domain of IE2; P. Robinson and T. Shenk, unpublished data), CREB/ATF-1 (p43, AB3006; Chemicon), or HCMV pUL24 (6D7) (12) used as a control antibody. The immunocomplexes were captured by the addition of 60 μl of protein G-agarose slurry (Upstate Biotechnology) for 2 h at 4°C with rotation. The agarose beads were collected by centrifugation (1,000 rpm for 1 min) and then washed for 5 min at 4°C with rotation sequentially with the following buffers: low-salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 8.0], 150 mM NaCl), high-salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl [pH 8.0], 500 mM NaCl), LiCl buffer (250 mM LiCl, 1% Nonidet P-40, 1% sodium deoxycholate, 1 mM EDTA, 10 mM Tris-HCl [pH 8.0]), TE buffer (1 mM EDTA, 10 mM Tris-HCl [pH 8.0]) and TE-NaCl buffer (50 mM NaCl, 1 mM EDTA, 10 mM Tris-HCl [pH 8.0]). The immunocomplexes were eluted from the beads by adding 100 μl of elution buffer (1% SDS, 100 mM NaHCO3), followed by incubation for 15 min at room temperature with rotation. After the beads were collected by centrifugation (1,000 rpm for 1 min), the supernatant was transferred to a new tube, and the elution was repeated. The supernatants were combined, NaCl was added to a final concentration of 200 mM, and the solution was incubated at 65°C for 5 h to reverse cross-linking. Proteins were digested by adding Tris-HCl (pH 6.5; final concentration, 40 mM), EDTA (final concentration, 10 mM), and proteinase K (final concentration, 40 μg/ml), followed by incubation for 2 h at 55°C, and the DNA was recovered by using QiaQuick purification columns (Qiagen) as specified by the manufacturer.
The ChIP experiments were analyzed by quantitative PCR (qPCR) by using a 7800HT sequence detector system (Applied Biosystems) with primer pairs specific for the immediate-early promoters MIEP (5′-AACAGCGTGGATGGCGTCTCC-3′ and 5′-GGCACCAAAATCAACGGGACTTT-3′) and UL37 (5′-GCGGGAGAGGATCTTCAAGG-3′ and 5′-CTCGGAAACTGTGCGTCAATG-3′), the early promoters UL112 (5′-GCTCCCAGCCTCTGTTAGGTTG-3′ and 5′-CATCATCTTTCCAGCCCGC-3′) and UL44 (5′-GCGTGCAAGTCTCGACTAAGGAGC-3′ and 5′-AAGTACTGTGCCTCTTAGTCGGGGG-3′), the late promoters UL94 (5′-CCGTCGAGTACGTGCTGATTCG-3′ and 5′-TGACGGCAAAGTTCCCAAACAAC-3′) and UL99 (5′-GAGGAAAGCGAACTGGGCTG-3′ and 5′-TCGTAGGAGCGTAGAGACACCTGG-3′), the viral nonpromoter region within the open reading frame of UL69 (5′-CTCGTCGTGTGACAGCAGGATG-3′ and 5′-GAACTACAGCAACTCAGCCGTTTGA-3′), and the intergenic region between TRL6 and TRL5 (5′-CAACGGTCGTCAATACAACAGCCT-3′ and 5′-CTCTCCACAGTTCACCATCTTCTTCG-3′). Primers for qPCR were verified to amplify a single amplicon. All qPCR amplifications were performed by using a Power Sybr green PCR kit (Applied Biosystems). The amplification conditions were performed at 50°C for 2 min, followed by 95°C for 10 min, and 40 cycles of 95°C for 15 s and 60°C for 1 min. The efficiency of amplification was determined for each primer pair by generating a standard curve with 10-fold serial dilutions of a known concentration of viral DNA. The slope values of the standard curve for the primer pair amplicons ranged from −3.5 to −3.2, indicating 90 to 100% efficiency. At the end of the qPCR run a dissociation curve was performed to ensure that each primer pair generated a single product of amplification, a process that was confirmed by agarose gel electrophoresis. For each primer pair, a no-template control was included, and each sample was run in duplicate.
In order to determine the amount of viral RNA in infected cells by qPCR, total RNA was isolated from fibroblasts at 6, 12, 24 and 48 hpi using the TRI-reagent (Sigma); treated with DNA-free reagents (Ambion) to remove contaminating DNA from the sample; and then used as a template for cDNA synthesis in a reaction that was primed with oligo(dT) according to the manufacturer's protocol (Applied Biosystems). The qPCR reaction was performed using a Sybr green PCR master mix (Applied Biosystems) and primers specific to UL122 (5′-ATGGTTTTGCAGGCTTTGATG-3′ and 5′-ACCTGCCCTTCACGATTCC-3′), UL37 (5′-GACGAAGTCCGATGAGGAGGATG-3′ and 5′-TGGGACACTGGGCGTTGTTG-3′), UL44 (5′-TACAACAGCGTGTCGTGCTCCG-3′ and 5′-GGCGTGAAAAACATGCGTATCAAC-3′), UL99 (5′-GTGTCCCATTCCCGACTCG-3′ and 5′-TTCACAACGTCCACCCACC-3′), or glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 5′-CTGTTGCTGTAGCCAAATTCGT-3′ and 5′-ACCCACTCCTCCACCTTTGAC-3′).
To analyze proteins by Western blot, infected cells were scraped into medium at 6, 12, 24, and 48 hpi and then pelleted by centrifugation (1,000 rpm for 5 min). After the pellet was washed with PBS, the cells were lysed in radioimmunoprecipitation assay buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM Tris-HCl [pH 8.0], protease inhibitors) and incubated on ice for 30 min. Cellular debris was removed by centrifugation (14,000 rpm for 10 min), and the protein concentration was determined with the Bio-Rad protein assay reagent (Bio-Rad). Equal amounts of the protein were mixed with SDS loading buffer (0.05 mM Tris-HCl [pH 6.8], 2% SDS, 10% glycerol, 0.1% bromophenol blue, 600 mM dithiothreitol) and subjected to electrophoresis in a 10% SDS-polyacrylamide gel. The resolved proteins were transferred to a nitrocellulose membrane, which was then incubated in blocking buffer (20 mM Tris-HCl [pH 7.4], 500 mM NaCl, 0.05% Tween 20, 5% nonfat milk). Proteins were detected using primary antibodies against IE2 (3A9 monoclonal antibody produced by immunization with a fusion protein containing the C-terminal domain of IE2; P. Robinson and T. Shenk, unpublished data), pUL44 (CA006-100; Virusys), pUL99 (10B4-29) (56), and tubulin (DM1A; Sigma) and then detected with a secondary goat anti-mouse antibody coupled to horseradish peroxidase (Amersham). The protein antibody complexes were visualized by treating the membrane with the ECL protein detection reagent (Amersham) and then exposing it to photographic film (Kodak).
Before we analyzed the acetylation patterns of histone H3 at the viral promoters, we performed a control ChIP assay to quantify the amount of DNA that is precipitated in the presence of a nonspecific antibody compared to no antibody. There was no significant difference in the two controls at the MIEP (Fig. (Fig.1A)1A) or at any of the other promoter and nonpromoter regions tested (data not shown). We also tested the specificity of the primer pairs we designed for the viral promoter and nonpromoter sequences by performing ChIP assays from mock-infected cells with antibodies specific for histone H3 acetylated at K9 or K14 (H3K9 and H3K14) and dimethylated histone H3K9 and comparing them to ChIP assays done with the same antibodies on infected cells harvested at 24 hpi. As expected, the qPCR signal from the mock-infected cells was dramatically less than that from infected cells (Fig. (Fig.1B),1B), and similar results were obtained for primers pairs from the other promoter and nonpromoter regions tested (data not shown). Having observed no difference between the ChIPs done with or without the control antibody and detecting minimal amplification when no viral DNA was used in the qPCR reaction, we normalized our ChIP results that follow to “no-antibody” controls, and we omitted the “no-template” control from the plots.
Since HCMV genes are expressed in a regulated cascade (42), we would expect that histone acetylation occurs at immediate-early promoters before early and late promoters. In order to quantify the relative levels of acetylated histone H3 at viral promoters, ChIP assays were performed at various times after infection using antibodies specific for acetylated H3K9 and H3K14. At 3 and 6 hpi the acetylation of H3K9 and H3K14 was pronounced at the immediate-early MIEP and UL37 promoter (Fig. 1C and G), except that the initial acetylation at H3K14 was not as robust as that observed for H3K9 at the UL37 promoter. At 12 hpi the acetylation of H3K9 at the MIEP and UL37 promoter decreased, but then it rose again at 24 hpi and at subsequent times; a similar pattern was observed for H3K14, except that the decrease at the UL37 promoter at 12 hpi is relatively modest. There was little H3K9 or H3K14 acetylation at early and late promoters during the first 12 h (Fig. 1D, E, H, and I). Once the immediate-early genes are expressed, they induce the activation of early genes that sponsor viral DNA replication and full induction of late promoters. Our results are consistent with this dynamic, since the ChIP assays documented an increase in H3K9 and H3K14 acetylation at early and late promoters starting at 24 hpi. The acetylation of H3 at DNA within the UL69 open reading frame and the intergenic region between the 3′ ends of TRL6 and TRL5, two regions that we presume do not include functional promoter elements, also increased as the infection progressed into the early and late phases (Fig. 1F and J). This suggests that increased acetylation at these times is not restricted to promoter regions but includes nonpromoter sequences as well.
We next sought to determine whether the higher level of H3 acetylation evident after 24 hpi was caused by increased acetylation of H3 or simply by an increase in the amount of H3 occupying the viral promoters. In order to address this, we performed ChIP assays at different time points, using an antibody specific for the carboxy-terminal domain of H3, thus recognizing it independently of its posttranslational modifications. The results in Fig. Fig.22 show that the viral genome has become associated with H3 at the earliest time tested (3 hpi). The variation in the amount of H3 associated with different segments of the viral genome that occurred over time did not mimic the increases and decreases in H3 modifications. Therefore, the increase in H3K9 and H3K14 acetylation detected in our experiments was due to hyperacetylation of the histone.
Dimethylation of H3 at lysine 9 is associated with silenced promoters (45, 59), and it has been reported that there is interplay between the acetylation and methylation of H3K9 leading to either active or inactive chromatin structures (47). Since early and late promoters did not become strongly acetylated until 24 hpi, we anticipated that chromatin at these promoters was methylated early in infection. Consistent with our prediction, the ChIP assays performed with an antibody specific for dimethyl H3K9 revealed substantial methylation of the early and late promoters at 3 and 6 hpi, and then methylation decreased as the infection progressed (Fig. 3B and C). In contrast, there was relatively little dimethyl H3K9 at immediate-early promoters during the first 24 h of infection, but methylation increased substantially at 48 and 72 hpi (Fig. (Fig.3A).3A). The increased dimethyl H3K9 at the immediate-early promoters at later times occurred in spite of the fact that the same promoters were found to contain acetylated H3K9 at these times (Fig. (Fig.1A).1A). These data suggest that there could be two populations of viral genomes late in infection, one with methylated H3K9 and another with acetylated H3K9. IE2 binds at the MIEP and blocks its activity at least in part by recruiting chromatin remodeling enzymes (8, 29, 30, 32, 36, 50). Perhaps as the number of genomes increases, there is not sufficient IE2 available to occupy all of the binding sites, and this could cause some H3K9 sites remain acetylated while others are methylated.
Previous studies have shown that in latently infected cells, HDAC inhibitors induce the expression of IE transcripts (44), and during lytic replication HDAC inhibitors restore the growth of an IE1-deficient virus to wild-type levels (46). These findings are consistent with the view that a block to HDAC activity favors viral gene transcription by increasing histone acetylation at viral promoters. To directly monitor the effect of HDAC inhibitors on the acetylation state of histone H3 at viral promoters, MRC-5 cells were treated with an HDAC inhibitor for 24 h prior to the infection. After infection, the cells were maintained in the presence of the drug and then processed for ChIP assays after various time intervals. An acetyl H3K9-specific antibody was used to probe chromatin from untreated cells (open bars) or cells treated with sodium butyrate (gray bars) or TSA (dark bars). The HDAC inhibitors induced an increase in H3K9 acetylation at all viral DNA regions studied at 6 and 12 hpi compared to the untreated cells (Fig. (Fig.4).4). At 24 hpi, HDAC inhibitors enhanced acetylation at most regions assayed; but at 48 hpi, H3K9 acetylation at some promoter regions was reduced by treatment with inhibitors. We do not have an explanation for this result, although it could be an indirect consequence of a drug effect at another locus influencing acetylation.
Since treatment with either HDAC inhibitor led to an increase in H3K9 acetylation of viral promoters during the immediate-early and early phases of the infection cycle, we anticipated that transcription of early and late genes might take place earlier in the infection. Accordingly, we tested the effect of the drugs on the accumulation of viral RNAs and proteins. HDAC inhibitors did not detectably alter the levels of IE2-86 or UL37 RNA at 6 hpi, but the immediate-early RNAs were elevated at 12 hpi and the level of UL37 RNA was markedly enhanced at later times (Fig. 5A and B). In contrast, we did not detect a difference in the accumulation of an early or late RNA (Fig. 5C and D). Consistent with the changes in RNA levels, the HDAC inhibitors induced a modest increase in the IE2 protein at 12 hpi and a significantly more substantial increase in UL37 protein, with little effect on early or late viral proteins (Fig. (Fig.5E).5E). The relatively muted increase in IE2 RNA and protein in response to drug treatment probably results from IE2 autoregulation (29, 32, 49). Nevertheless, the enhanced accumulation of immediate-early products at 12 hpi is consistent with the ability of HDAC inhibitors to enhance H3K9 acetylation at that time.
We tested whether the acetylation of viral chromatin is affected by the expression of viral genes and replication of the viral genome. We treated infected MRC-5 cells with cycloheximide to block translation or PAA to block viral DNA replication beginning 1 h prior to infection. Cells were harvested at various times after infection, and H3K9 acetylation of viral chromatin was monitored by ChIP. In untreated cells, MIEP acetylation again decreased from 6 to 12 hpi and then increased at 24 and 48 hpi, but cycloheximide or PAA prevented the increase at later times after infection (Fig. (Fig.6A).6A). The immediate-early UL37 promoter exhibited a similar H3K9 acetylation pattern as observed for the MIEP (Fig. (Fig.6B),6B), but acetylation of early and late promoters was substantially reduced by PAA and almost completely blocked by cycloheximide treatment at all times tested (Fig. 6C and D). The drug treatments caused predictable changes in viral RNA and protein accumulation (data not shown). Viral DNA replication must take place in order for immediate-early promoters to be active during the late phase of infection, since acetylation at the MIEP and UL37 promoter (Fig. 6A and B), as well as the accumulation of their corresponding RNA and protein (data not shown) did not increase after 12 hpi in the presence of PAA.
It is known that IE2 can repress the activity of the MIEP and US3 promoter by binding to their CRS motifs (30, 32, 49), as well as activate other early and late promoters (26, 37). It has been shown that IE2 can bind the MIEP, UL112, and UL4 promoters by performing ChIP assays after transfecting cells with vectors expressing wild-type and mutant forms of IE2 together with plasmids containing the viral promoter sequences (48). We wanted to extend these earlier observations by using ChIP assays to test for the presence of IE2 at different viral promoters as a function of time within the context of infected cells. As expected, IE2 was found at the MIEP. None was detected at 3 hpi, intermediate levels were present at 6 and 12 hpi, and the highest levels of IE2 were evident at 24, 48, and 72 hpi (Fig. (Fig.7A).7A). In contrast, the UL37 immediate-early promoter did not exhibit detectable IE2 association (Fig. (Fig.7A).7A). This promoter in not known to contain an IE2 binding motif and, like other immediate-early promoters, it does not require IE2 for its activation. IE2 accumulated as the infection progressed at the two early and two late promoters tested (Fig. 7B and C). Of these promoters, immunoprecipitation of viral DNA is most efficient for UL112, which contains a binding site for IE2 (1), the other early and late promoters that were assayed are not known to contain a binding site for IE2. The viral protein could become associated with these promoters by interacting with other DNA-binding factors. Figure Figure7D7D shows that IE2 could not be detected in association with two presumptive nonpromoter regions of the HCMV genome.
CREB/ATF is a cellular transcription factor that associates with IE2 to induce transcription (28), and it is important for activity of the UL112 early promoter (53, 55). CREB/ATF is also likely to regulate the MIEP since it contains four CREB/ATF binding sites (19). Consequently, we tested whether CREB/ATF could be found at viral promoters by ChIP assay. An interaction of CREB/ATF was evident at the MIEP and UL112 early promoter, and marginal interactions were evident at 24 to 72 hpi for the other immediate-early and early promoters (Fig. 8A and B). No interaction was evident at late promoters or in nonpromoter regions (Fig. 8C and D).
The acetylation of H3K9 and H3K14 at the MIEP was initially high but then decreased at 12 hpi, before increasing again at later times (Fig. 1C and G). It seemed likely that the transient decrease could be caused by IE2 binding at the CRS within the MIEP. IE2 inhibits the MIEP (32, 49, 50), at least in part by recruiting proteins with HDAC and methyltransferase activity (32, 49, 50). In order to test whether binding of IE2 could be causing the decreased acetylation at the MIEP, we generated a virus (BADsubUL122crs−) containing a 3-bp substitution mutation within the CRS, which has been shown to abrogate IE2 binding (35). The mutant grew in fibroblasts, but produced an ~200-fold reduced yield (data not shown). We infected fibroblasts with the mutant or wild-type virus at equal multiplicities (equal particles, determined by qPCR), and cells were harvested 6, 12, 24 and 48 h later. As expected, there was no difference in the amount of MIEP-specific DNA precipitated when a nonspecific antibody was compared to a control with no antibody (Fig. (Fig.9A).9A). A second set of control ChIP assays confirmed that IE2 binding at the MIEP containing the mutant CRS was substantially reduced (Fig. (Fig.9B).9B). The small amount of IE2 detected at the promoter could result from some residual binding at the mutated element, the presence of another IE2 binding site in the MIEP (18), or recruitment of IE2 through an interaction with another protein at the MIEP. The amount of IE2 at the UL112 promoter was not affected by the mutation in the MIEP, and the UL37 promoter served as a negative control (Fig. (Fig.9B).9B). Having shown that binding of IE2 to the MIEP is decreased in the mutant virus, we next performed a time course experiment to analyze the acetylation of H3K9 at the MIEP. The results in Fig. Fig.9C9C show that, in contrast to the wild-type virus, acetylation of H3 at the MIEP did not decrease at 12 hpi with the mutant lacking a functional CRS. Further, at 24 and 48 hpi acetylation of H3 is greater at the mutant MIEP than at the wild-type MIEP. Since IE2 is capable of repressing the activity of the MIEP, our results argue that the increased acetylation of the MIEP at 12, 24 and 48 hpi with the mutant virus is likely caused by inefficient IE2 binding at the MIEP.
Although the total amount of histone H3 associated with the viral genome remained fairly constant (Fig. (Fig.2),2), we observed dynamic acetylation and methylation patterns of HCMV chromatin at different stages of the infectious cycle. The patterns fall into two groupings: immediate-early promoters versus all of the other promoters and nonpromoter regions (Fig. (Fig.11 and and3).3). At the start of infection, histone H3 at immediate-early promoters was hypomethylated (K9) and hyperacetylated (K9 and K14). In contrast, early and late promoters, as well as nonpromoter domains, contained hyper-methylated and hypo-acetylated H3 at 3 and 6 hpi. As the infection entered the late phase, the level of H3 methylation at the immediate-early promoter increased, whereas H3 methylation at the other promoters and nonpromoter regions was reduced. Late after infection all regions of the viral genome tested were associated with hyper-acetylated H3. The late hyper-acetylation of H3 was substantially blocked by cycloheximide or PAA treatment (Fig. (Fig.6),6), indicating that the late modification requires continuing cellular and/or viral gene expression and viral DNA replication.
The key to the different patterns of methylation and acetylation at immediate-early compared to early and late promoters probably resides in their organization. Early and late promoters are relatively simple in terms or their constituent transcription factor binding sites, whereas immediate-early promoters are more complex. For instance, the MIEP promoter/enhancer region includes five CREB/ATF 19-bp repeat sites, four NF-κB 18-bp repeat sequences, three 21-bp repeats with Sp1, YY1 and ERF binding sites, three retinoic acid receptor response elements and two AP1 consensus sites (40). Although individual binding sites or groups of sites, such as the full set of NF-κB (5) or CREB/ATF (25) sites, can be deleted with little apparent effect on function of the MIEP during viral replication in fibroblasts, mutation of Sp1 sites (22) and deletion or substitution of substantial portions of the enhancer region reduce its activity (21, 39). Some of the cellular factors binding at the MIEP almost certainly direct histone modifying activities to the promoter/enhancer at the start of infection. For example, YY1 is known to bind the histone acetyltransferases and HDACs (62), as well as histone methyltransferases (4, 52).
The level of acetylated H3 at immediate-early promoters is initially higher (3 and 6 hpi) before it is reduced at 12 hpi. We believe that this dynamic behavior is mediated by the viral IE2 protein. Relatively little IE2 is present at the start of the infection, but as it accumulates it binds to the CRS within the MIEP (1, 8, 24, 29, 49) and recruits chromatin remodeling factors that silence transcription (50). Consistent with this interpretation, the reduced acetylation of histone H3 at 12 hpi does not occur after infection with a mutant virus lacking a functional CRS, and MIEP H3 acetylation at 24 and 48 hpi is greater than in the wild-type virus (Fig. (Fig.9).9). HDAC inhibitors also blocked the transient hypoacetylation at 12 hpi (Fig. (Fig.4)4) with an increase in IE2 expression at this time (Fig. (Fig.5).5). Inhibition of protein synthesis by treatment with cycloheximide also modulated acetylation at immediate-early promoters, inducing an increase in acetylation of H3 (Fig. (Fig.6)6) with increased expression of immediate-early RNAs (data not shown). Since the drug was added before the start of infection, precluding the synthesis of IE2, part of the cycloheximide effect could result from the failure of IE2 to bind the MIEP and block its activity. However, the UL37 immediate-early promoter also undergoes transient hypoacetylation at 12 hpi (Fig. (Fig.4),4), and was hyperacetylated in response to cycloheximide (Fig. (Fig.6).6). We did not detect significant levels of IE2 at this promoter (Fig. (Fig.7),7), arguing that additional mechanisms contribute to the changes in histone acetylation and elevated activity in the absence of protein synthesis. It seems likely that the cycloheximide effect on acetylation is due in part to differential stabilities of proteins involved in competing acetylation and deacetylation activities at the promoters. Possibly, a critical HDAC or factor that directs it to viral promoters has a shorter half-life than the activities promoting acetylation.
Why is the IE2-controlled hypoacetylation at the MIEP transient? It is likely that the dynamic effect results from the relative numbers of viral genomes versus IE2 molecules. Initially (3 and 6 hpi), IE2 has not accumulated to significant levels, most copies of the MIEP CRS are not occupied, histones at the MIEP are hyperacetylated and the promoter is highly active. At 12 hpi the ratio of IE2 to genomes is relatively high, because the MIEP has been maximally active, and its product, IE2, has accumulated, but the genome copy number has remained stable because DNA replication has not yet begun. As a consequence, IE2 occupies the CRS, recruits HDACs and inhibits MIEP activity. Then, less IE2 is produced as DNA replication begins and the genome copy number increases. At this point there is no longer sufficient IE2 available to occupy the available CRS motifs, and many copies of the promoter are again acetylated and highly active.
In contrast to its inhibitory behavior at the MIEP, IE2 can activate cellular and viral genes by recruiting histone acetyltransferases to cellular and viral promoters to initiate transcription (6). Consistent with this earlier work, we detected IE2 at multiple early and late promoters after 24 hpi (Fig. (Fig.7),7), a time when promoters were active (Fig. (Fig.5)5) and contained hyperacetylated H3 (Fig. (Fig.1).1). Our results reinforce the view that the ability of IE2 to bind to histone acetyltransferases contributes importantly to its transcriptional regulatory activities (38, 64). Importantly, IE2 is not the only factor that leads to increased H3 acetylation on HCMV chromatin, since nonpromoter regions of the genome where IE2 does not bind at a detectable level (Fig. (Fig.7)7) are also acetylated after 24 hpi. It appears that replication of viral DNA is another important event that causes the acetylation of the viral genome, since treatment with PAA to block viral DNA replication caused hypoacetylation of H3 at viral promoters and nonpromoter regions (Fig. (Fig.4).4). It is not clear why nonpromoter regions are acetylated. Perhaps the H3 acetylation late after infection is nonspecific and not functionally relevant. Alternatively, the modification might play a role in other events in the viral replication cycle such as DNA synthesis and packaging.
We tested the association of CREB/ATF with viral promoters because this cellular transcription factor has been shown to bind IE2 (28). Even though IE2 was present at promoters other than UL37 (Fig. (Fig.7),7), CREB/ATF was detected at only the MIEP and at the early UL112 promoter (Fig. (Fig.8).8). As noted above, the activity of the MIEP within infected fibroblasts is unaffected when all five CREB/ATF-binding sites are deleted (25), but it is likely that CREB/ATF will prove important for the MIEP function in different cell types or under different physiological conditions that have not yet been tested. CREB/ATF is important for efficient UL112 transcription (53, 55), where it might interact with DNA-bound IE2.
Our observations suggest that the virus exploits a competition between chromatin remodeling factors that acetylate histones to induce viral gene activation and factors that methylate histones to silence viral transcription. This competition is likely part of the mechanism by which the virus expresses its genes in an ordered cascade (Fig. (Fig.10).10). Once the viral genome enters the nucleus and associates with histones, the H3 at immediate-early promoters is hyperacetylated and hypomethylated, arguing that they successfully recruit transcription factors, histone acetyltransferases, as well as other components of the basal transcriptional machinery components needed to initiate transcription. In contrast, early and late promoters are bound by hypoacetylated and hypermethylated H3, arguing that they either fail to recruit histone acetyltransferases or recruit HDACs, DNA, and histone methyltransferases. As viral immediate-early proteins accumulate, IE1 favors more global acetylation of the viral chromatin by blocking the activity of HDACs (46) and IE2 recruits histone acetyltransferases (6) to induce the activity of early and late promoters. As the infection progresses through the late phase (after 48 hpi), new virions are assembled. At this late time, the MIEP is bound to both acetylated (Fig. (Fig.1)1) and methylated histone H3 (Fig. (Fig.3).3). Perhaps these differentially modified histones reside at the same promoter. Alternatively, they could represent two distinct populations of HCMV chromosomes (Fig. 10B). The population with methylated H3 might be transcriptionally inactive because of IE2 binding to the CRS, whereas the population with acetylated H3 might be actively transcribed. The failure to deacetylate histones at the MIEP could result from limited steady-state levels of IE2 late after infection, and it could conceivably mark viral genomes for packaging.
It is well established that DNA methylation, histone deacetylation, methylation of histone H3 at lysine 9 and its association with HP1 act as epigenetic markers to silence gene expression. Further, it is clear that factors controlling DNA and histone methylation physically and functionally interact (13). What is not fully understood is the order in which the DNA and histone modifications take place, their interdependence, and how that interindependence varies among different loci. Experiments showing that a DNA methyltransferase complex recruits HDACs support a model in which DNA methylation leads to histone methylation (10, 54, 65). On the other hand, methylated H3K9 has been shown to act as a marker that leads to DNA methylation (14), and cells lacking H3K9 methyltransferases exhibit significantly diminished CpG methylation at some loci (23, 31). Regardless of the order in which DNA methylation, histone H3 methylation, and recruitment of HP1 occur, all of these combined processes are hallmarks of heterochromatin. Although the lack of methylation in bacterial DNA activates innate immunity (2), it has been proposed that DNA methylation might serve as a cellular defense against infection by silencing the genome of an invading virus (11). It is conceivable that the histone methylation we observed early after infection reflects an innate response to viral DNA.
This study was supported by a grant from the National Institutes of Health (CA85786).
Published ahead of print on 23 July 2008.